Global tests of P-values for multifactor dimensionality reduction models in selection of optimal number of target genes

Dai, Hongying; Bhandary, Madhusudan; Becker, Mara; Leeder, J Steven; Gaedigk, Roger; Motsinger-Reif, Alison A

doi:10.1186/1756-0381-5-3

BioData Mining

Table 6 Power of six global tests of correlated P-values (The correlation coefficient for Beta random variables is ρ Uniform distributions have random correlation matrices)

From: Global tests of P-values for multifactor dimensionality reduction models in selection of optimal number of target genes

Correlated 0.9 Uniform (0,1) ± 0.1Beta (0.4,6), $ρ = 0.8$
n	KS	Inverse chi	Inverse norm	Tippett	Wilcoxon	Logit
20	0.19	0.402	0.285	0.329	0.204	0.323
50	0.293	0.654	0.458	0.547	0.324	0.514
100	0.41	0.833	0.644	0.665	0.48	0.7
200	0.705	0.963	0.856	0.785	0.693	0.901
300	0.864	0.989	0.943	0.861	0.834	0.956
400	0.934	0.996	0.967	0.878	0.862	0.978
500	0.962	0.999	0.989	0.914	0.928	0.994
Correlated 0.6 Uniform (0,1) ± 0.4Beta (0.4,6), $ρ = 0.8$
n	KS	Inverse chi	Inverse norm	Tippett	Wilcoxon	Logit
20	0.848	0.968	0.913	0.787	0.837	0.924
50	0.994	0.999	0.998	0.922	0.989	0.998
100	1	1	0.999	0.956	0.999	0.999
200	1	1	1	0.987	1	1
300	1	1	1	0.993	1	1
400	1	1	1	0.992	1	1
500	1	1	1	0.994	1	1
Correlated 0.9 Uniform (0,1) ± 0.1Beta (0.5,4.5), $ρ = 0.8$
n	KS	Inverse chi	Inverse norm	Tippett	Wilcoxon	Logit
20	0.145	0.282	0.199	0.243	0.158	0.237
50	0.281	0.491	0.376	0.345	0.296	0.409
100	0.374	0.686	0.532	0.425	0.418	0.573
200	0.598	0.874	0.734	0.531	0.627	0.771
300	0.774	0.951	0.866	0.579	0.767	0.89
400	0.855	0.975	0.914	0.641	0.836	0.936
500	0.938	0.993	0.962	0.635	0.91	0.977
Correlated 0.6 Uniform (0,1) ± 0.4Beta (0.5,4.5), $ρ = 0.8$
n	KS	Inverse chi	Inverse norm	Tippett	Wilcoxon	Logit
20	0.149	0.287	0.221	0.217	0.178	0.242
50	0.265	0.518	0.4	0.334	0.323	0.438
100	0.421	0.693	0.539	0.434	0.438	0.575
200	0.604	0.856	0.726	0.485	0.644	0.754
300	0.778	0.951	0.858	0.582	0.761	0.886
400	0.871	0.978	0.931	0.612	0.858	0.942
500	0.921	0.981	0.948	0.658	0.901	0.957
Correlated 0.9 Uniform (0,1) ± 0.1Beta (1,5), $ρ = 0.8$
n	KS	Inverse chi	Inverse norm	Tippett	Wilcoxon	Logit
20	0.157	0.147	0.158	0.066	0.154	0.154
50	0.192	0.234	0.231	0.069	0.222	0.22
100	0.32	0.355	0.347	0.064	0.344	0.34
200	0.491	0.521	0.501	0.061	0.502	0.488
300	0.655	0.698	0.663	0.068	0.651	0.652
400	0.776	0.784	0.754	0.068	0.751	0.741
500	0.847	0.844	0.82	0.072	0.823	0.802
Correlated 0.6 Uniform(0,1) ± 0.4Beta(1,5), $ρ = 0.8$
n	KS	Inverse chi	Inverse norm	Tippett	Wilcoxon	Logit
20	0.698	0.676	0.695	0.124	0.703	0.682
50	0.966	0.921	0.938	0.131	0.948	0.922
100	0.999	0.988	0.991	0.11	0.997	0.988
200	1	1	1	0.116	1	1
300	1	1	1	0.122	1	1
400	1	1	1	0.122	1	1
500	1	1	1	0.114	1	1
Correlated 0.9 Uniform(0,1) ± 0.1Beta(0.4,6), $ρ = Beta (2, 5)$
n	KS	Inverse chi	Inverse norm	Tippett	Wilcoxon	Logit
20	0.147	0.349	0.245	0.332	0.181	0.288
50	0.263	0.59	0.422	0.529	0.298	0.481
100	0.396	0.804	0.627	0.648	0.447	0.687
200	0.675	0.963	0.831	0.804	0.662	0.89
300	0.809	0.99	0.927	0.836	0.808	0.951
400	0.899	0.998	0.966	0.907	0.865	0.979
500	0.958	0.999	0.98	0.928	0.912	0.99
Correlated 0.6 Uniform(0,1) ± 0.4Beta(0.4,6), $ρ = Beta (2, 5)$
n	KS	Inverse chi	Inverse norm	Tippett	Wilcoxon	Logit
20	0.827	0.974	0.921	0.789	0.808	0.941
50	0.994	1	0.999	0.945	0.977	0.999
100	1	1	1	0.983	1	1
200	1	1	1	0.997	1	1
300	1	1	1	1	1	1
400	1	1	1	1	1	1
500	1	1	1	0.999	1	1
Correlated 0.9 Uniform(0,1) ± 0.1Beta(0.5,4.5), $ρ = Beta (2, 5)$
n	KS	Inverse chi	Inverse norm	Tippett	Wilcoxon	Logit
20	0.141	0.269	0.205	0.229	0.162	0.231
50	0.235	0.444	0.33	0.321	0.257	0.367
100	0.359	0.661	0.526	0.408	0.423	0.562
200	0.543	0.879	0.709	0.516	0.577	0.747
300	0.726	0.95	0.849	0.562	0.731	0.878
400	0.81	0.977	0.902	0.632	0.81	0.928
500	0.914	0.993	0.959	0.683	0.895	0.967
Correlated 0.6 Uniform (0,1) ± 0.4Beta (0.5,4.5), $ρ = Beta (2, 5)$
n	KS	Inverse chi	Inverse norm	Tippett	Wilcoxon	Logit
20	0.138	0.26	0.199	0.238	0.169	0.212
50	0.241	0.469	0.356	0.335	0.288	0.385
100	0.347	0.662	0.506	0.402	0.408	0.55
200	0.547	0.885	0.703	0.519	0.585	0.761
300	0.694	0.949	0.852	0.592	0.727	0.88
400	0.802	0.987	0.914	0.627	0.804	0.935
500	0.882	0.991	0.948	0.641	0.871	0.959
Correlated 0.9 Uniform(0,1) ± 0.1Beta(1,5), $ρ = Beta (2, 5)$
n	KS	Inverse chi	Inverse norm	Tippett	Wilcoxon	Logit
20	0.121	0.118	0.133	0.071	0.132	0.129
50	0.194	0.21	0.214	0.064	0.217	0.211
100	0.277	0.311	0.305	0.071	0.306	0.299
200	0.449	0.502	0.486	0.068	0.485	0.477
300	0.585	0.643	0.612	0.058	0.623	0.599
400	0.66	0.72	0.697	0.075	0.696	0.682
500	0.767	0.8	0.775	0.059	0.759	0.759
Correlated 0.6 Uniform(0,1) ± 0.4Beta(1,5), $ρ = Beta (2, 5)$
n	KS	Inverse chi	Inverse norm	Tippett	Wilcoxon	Logit
20	0.594	0.591	0.632	0.124	0.637	0.616
50	0.913	0.9	0.921	0.139	0.922	0.914
100	0.996	0.992	0.992	0.111	0.995	0.99
200	1	1	1	0.102	1	1
300	1	1	1	0.115	1	1
400	1	1	1	0.116	1	1
500	1	1	1	0.095	1	1
Correlated 0.9 Uniform(0,1) ± 0.1Beta(0.4,6), $ρ = Uniform (0.1, 0.9)$
n	KS	Inverse chi	Inverse norm	Tippett	Wilcoxon	Logit
20	0.183	0.397	0.292	0.363	0.217	0.324
50	0.291	0.639	0.459	0.545	0.34	0.508
100	0.418	0.821	0.625	0.653	0.463	0.694
200	0.669	0.958	0.855	0.78	0.68	0.881
300	0.818	0.993	0.924	0.861	0.782	0.947
400	0.919	0.997	0.97	0.908	0.89	0.985
500	0.974	1	0.988	0.926	0.932	0.994
Correlated 0.6 Uniform(0,1) ± 0.4Beta(0.4,6), $ρ = Uniform (0.1, 0.9)$
n	KS	Inverse chi	Inverse norm	Tippett	Wilcoxon	Logit
20	0.815	0.964	0.928	0.825	0.836	0.945
50	0.99	1	0.997	0.92	0.988	0.997
100	1	1	1	0.971	1	1
200	1	1	1	0.993	1	1
300	1	1	1	0.997	1	1
400	1	1	1	0.998	1	1
500	1	1	1	1	1	1
Correlated 0.9 Uniform(0,1) ± 0.1Beta(0.5,4.5), $ρ = Uniform (0.1, 0.9)$
n	KS	Inverse chi	Inverse norm	Tippett	Wilcoxon	Logit
20	0.154	0.283	0.216	0.228	0.171	0.24
50	0.272	0.465	0.352	0.331	0.296	0.391
100	0.358	0.675	0.512	0.421	0.414	0.561
200	0.528	0.871	0.706	0.526	0.578	0.749
300	0.749	0.958	0.857	0.578	0.752	0.89
400	0.857	0.986	0.924	0.632	0.841	0.938
500	0.906	0.99	0.951	0.679	0.877	0.962
Correlated 0.6 Uniform(0,1) ± 0.4Beta(0.5,4.5), $ρ = Uniform (0.1, 0.9)$
n	KS	Inverse chi	Inverse norm	Tippett	Wilcoxon	Logit
20	0.137	0.282	0.2	0.228	0.159	0.234
50	0.237	0.484	0.359	0.306	0.293	0.392
100	0.345	0.679	0.503	0.411	0.416	0.544
200	0.577	0.858	0.719	0.512	0.6	0.757
300	0.725	0.946	0.848	0.584	0.743	0.873
400	0.868	0.981	0.92	0.621	0.85	0.931
500	0.9	0.989	0.947	0.676	0.869	0.959
Correlated 0.9 Uniform(0,1) ± 0.1Beta(1,5), $ρ = Uniform (0.1, 0.9)$
n	KS	Inverse chi	Inverse norm	Tippett	Wilcoxon	Logit
20	0.141	0.166	0.16	0.078	0.155	0.155
50	0.213	0.224	0.235	0.054	0.251	0.232
100	0.318	0.362	0.364	0.073	0.365	0.362
200	0.477	0.525	0.498	0.047	0.503	0.485
300	0.621	0.651	0.637	0.071	0.622	0.621
400	0.718	0.747	0.721	0.071	0.708	0.707
500	0.804	0.818	0.794	0.075	0.785	0.78
Correlated 0.6 Uniform(0,1) ± 0.4Beta(1,5), $ρ = Uniform (0.1, 0.9)$
n	KS	Inverse chi	Inverse norm	Tippett	Wilcoxon	Logit
20	0.694	0.656	0.705	0.109	0.721	0.685
50	0.938	0.901	0.929	0.125	0.94	0.915
100	0.994	0.989	0.991	0.114	0.995	0.99
200	1	1	1	0.143	1	1
300	1	0.999	1	0.113	1	1
400	1	1	1	0.122	1	1
500	1	1	1	0.119	1	1

Uniform distributions have random correlation matrices).

Back to article page

ISSN: 1756-0381

Contact us

General enquiries: journalsubmissions@springernature.com